adaptive training #3

it's amazing what changing the targets can do. now they're spread evenly from the starting position (uniform distance) and they seem to do much better, rather than finding a local maximum (a single target that's easier to get to than the others).
run length:27
012345678910111213141516171819202122232425
avg0.12991110.15207480.20727720.31870840.39584120.64687490.60108480.76554070.92835640.95608071.270251.2402941.0785981.2572121.0924871.3274671.088351.1549031.3703621.1982141.2119511.1785781.2537440.95138751.3470260.8872428
max0.61878870.76564751.1996931.0606281.6245532.0148892.002721.5329191.7799932.9245082.0194351.8904252.370942.0550641.8125472.1251032.0440511.911442.0505212.573581.4987481.4173371.9088071.2616831.808071.39744
Population Size
100
Run Length
20
Do Training
False
Do Hybrid Training
True
Do Competitive Run
False
Do Adaptive Training
True
Inputs
hasTarget,targetLeft,targetRight,targetNorth,targetSouth,dirToTarget,distToTarget,wall,sensor0,sensor1,sensor2
Outputs
left,right,up,down,run,sensordir0,sensordir1,sensordir2
Hidden Layers
8 12 12 12 8
Back Propogation
True
Learning rate
0.5
Momentum
0.1
Growth rate
0.5